← Back to community benchmarks

Qwen3.5-4B

M2 Max (30c) · 32 GB · 4bit · 2026-03-10
Performance
4k
tokens
586.7
PP tok/s
91.0
TG tok/s
6981
TTFT (ms)
4.2
Peak mem (GB)
Hardware
Chip M2 Max (30c)
Memory 32 GB
GPU Cores 30
Software
oMLX v0.2.6
macOS macOS 15.7.3
Context 4,096
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 588.7 98.5 4.0 GB view
4k 586.7 91.0 4.2 GB current
Batching Results
Batch Size TG tok/s Speedup
98.5 1.00×
117.3 1.19×
131.6 1.34×